JAMM's Output

JAMM will create two folders in the output folder: "peaks" and "xcorr"

• "xcorr" contains results from fragment length estimation analysis if -f was not set by the user.
• "peaks" contains two files: all.peaks.narrowPeak and filtered.peaks.narrowPeak.

Peaks are given in ENCODE narrowPeak format. Note that JAMM's peak scores are in the 7th column. To obtain average enrichment of the peaks (signal / smoothed read count), divide the 7th column by the 9th column.

Peaks reported in the filtered list are a subset of the peaks reported in the all list. It is recommended to use the filtered.peaks.narrowPeak list. The peaks in the all.peaks.narowPeak file have a lot of "artefact" peaks (like single basepair peaks or peaks with very few reads). These "artefact" peaks result from the clustering procedure, (which is explained in the paper), and will always have very low scores and appear to the bottom of the list. The filtered.peaks.narrowPeak list will never contain any of these "artefact" peaks and this is why it is recommended to use the filtered.peaks.narrowPeak list. You can use the all.peaks.narrowPeak, if you prefer to do your own filtering, but in this case please be aware of this issues.

Hint: If you are directing stderr and stdout to a file, pipe them to the same file. The output will be more meaningful.